Amazon.com Inc’s cloud-computing unit on Tuesday announced updated versions of its in-house computer chips while also forging closer ties with Nvidia Corp — dual efforts designed to ensure it can get enough supplies of crucial data-center processors.
New homegrown Graviton4 chips will have as much as 30 percent better performance than their predecessors, Amazon Web Services (AWS) said at its annual re:Invent conference in Las Vegas. Computers using the processors are to start coming online in the coming months.
The company also unveiled Trainium2, an updated version of a processor designed for artificial intelligence (AI) systems. It will begin powering new services starting next year, Amazon said. That chip provides an alternative to so-called AI accelerators sold by Nvidia — processors that have been vital to the build-out of AI services.
Photo: Bloomberg
However, Amazon also touted “an expansion of its partnership” with Nvidia, whose CEO, Jensen Huang (黃仁勳), joined AWS CEO Adam Selipsky on stage.
AWS is to be the first big user of an updated version of that company’s Grace Hopper Superchip, and it will be one of the data-center companies hosting Nvidia’s DGX Cloud service.
The new in-house chips are part of AWS’s push to maintain its lead over Microsoft Corp’s Azure and Alphabet Inc’s Google Cloud Platform. Earlier this month, Microsoft announced its own processors and AI accelerators. Like Amazon, it has pointed to the efficiency that they bring compared with buying off-the-shelf parts from traditional sellers such as Intel Corp.
AWS said that it has built more than 2 million Graviton processors since beginning the project about five years ago. All of the top 100 of its users of EC2 — AWS’ family of processing power for rent — have chosen to use Graviton-based computing, the company said.
“By focusing our chip designs on real workloads that matter to customers, we’re able to deliver the most advanced cloud infrastructure to them,” AWS vice president Dave Brown said. “Graviton4 marks the fourth generation we’ve delivered in just five years. And with the surge of interest in generative AI, Trainium2 will help customers train their machine learning models faster, at a lower cost and with better energy efficiency.”
Amazon faces a delicate balancing act. Although it is keen to tout the technological prowess of its new chips, the company also needs to maintain its relationship with Nvidia. Getting sufficient supply of Nvidia products has become a status symbol in the technology industry, with figures such as Larry Ellison and Elon Musk boasting about their ability to get the chips.
For now, Nvidia’s products are considered the industry’s best. Until Trainium or other alternatives can match their capabilities, companies like Amazon cannot afford to alienate Nvidia — and catching up will not be easy. Nvidia’s prized H100 chip is getting a more advanced update called H200 in the first half of next year, followed by a whole new design later in the year.
AWS will also operate a supercomputer based on Nvidia hardware. When not in use for internal research and development, the “supercluster” will be made available to other users as a service.
Taiwan Semiconductor Manufacturing Co (TSMC, 台積電) has approved a capital budget of US$31.28 billion for production expansion to meet long-term development needs during the artificial intelligence (AI) boom. The company’s board meeting yesterday approved the capital appropriation plan for purposes such as the installation of advanced technology capacity and fab construction, the world’s largest contract chipmaker said in a statement. At an earnings conference last month, TSMC forecast that its capital expenditure for this year would be at the higher end of the US$52 billion to US$56 billion range it forecast in January in response to robust demand for 5G, AI and
NEW MARKET: The partnership opens up India to the Dutch company, which already has a strong hold in the semiconductor market of South Korea, Taiwan and China ASML Holding NV entered into a partnership agreement with Tata Electronics Pvt Ltd aimed at ramping up India’s goal to develop domestic chip-manufacturing capabilities. The Dutch company’s technology would help power Tata Electronics’ planned 300 millimeter (mm) semiconductor foundry in Gujarat, according to a joint statement from the two companies on Saturday. The signing of a memorandum of understanding coincides with a visit by Indian Prime Minister Narendra Modi to the Netherlands, which is looking to deepen bilateral relations with New Delhi. ASML, whose top customers include Taiwan Semiconductor Manufacturing Co (台積電) and Samsung Electronics Co, makes lithography machines that can print
PORTFOLIO REBALANCING: The adjustments in three global equity indices reflect rising investor appetite for semiconductor and artificial intelligence-related stocks Taiwan’s weighting in major global equity indices compiled by MSCI Inc is to rise modestly following the latest quarterly review, underscoring the market’s expanding role in emerging-market portfolios, as global investors continue to favor the nation’s technology sector. Taiwan’s weighting in the MSCI Emerging Markets Index is to increase by 0.30 percentage points to 23.76 percent, after the changes take effect at the close of the May 29 session. Its weighting in the MSCI All-Country Asia ex-Japan Index is to rise 0.37 percentage points to 27.16 percent, while that in the MSCI All Country World Index is to edge up slightly to
The Hsinchu County Government’s Labor Affairs Department yesterday said that it has received a plan from cosmetics brand Taiwan Shiseido Co (台灣資生堂) detailing mass layoffs at its plant in Hukou Township (湖口). While the labor authorities did not disclose the number of employees to be laid off, Japanese news media earlier in the day reported that the closure of the company’s factory in Hukou would result in 170 employees losing their jobs. Shiseido followed the law by reporting its layoff plan, the department said, adding that authorities would closely monitor negotiations between the management and affected employees and step in if any